Human Reproduction Update Advance Access originally published online on May 2, 2006
Human Reproduction Update 2006 12(4):351-361; doi:10.1093/humupd/dml017
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
The origins and sequelae of abnormal neuroendocrine function in polycystic ovary syndrome
1 The Center for Research in Reproduction and 2 Division of Endocrinology, Department of Internal Medicine, University of Virginia Health System, Charlottesville, VA, USA
3 To whom correspondence should be addressed at: Center for Research in Reproduction, Box 800391, University of Virginia Health System, Charlottesville, VA 22908, USA. E-mail: sek2h{at}virginia.edu
Submitted on December 13, 2005; resubmitted on March 10, 2006; accepted on March 27, 2006
 |
Abstract |
|---|
Polycystic ovary syndrome (PCOS) is a common clinical disorder characterized by ovulatory dysfunction and hyperandrogenaemia. A neuroendocrine hallmark of PCOS is persistently rapid LH (GnRH) pulsatility, which favours pituitary synthesis of LH over that of FSH and contributes to the increased LH concentrations and LH : FSH ratios typical of PCOS. Inadequate FSH levels contribute to impaired follicular development, whereas elevated LH levels augment ovarian androgen production. Whereas luteal phase elevations in progesterone normally slow GnRH pulse frequency, women with PCOS do not experience normal progesterone-mediated slowing, due in part to impaired hypothalamic progesterone sensitivity. This reduction in hypothalamic progesterone sensitivity appears to be mediated by elevated androgens because sensitivity can be restored with the androgen receptor blocker flutamide. The ovulatory and hormonal abnormalities associated with PCOS generally present during puberty, typically associated with hyperandrogenaemia. Along with elevated LH concentration and pulsatility, some girls with hyperandrogenaemia have impaired hypothalamic progesterone sensitivity similar to that seen in adult women with PCOS. We propose that peripubertal hyperandrogenaemia may lead to persistently rapid GnRH pulse frequency via impaired hypothalamic feedback inhibition. The subsequent abnormalities in gonadotropin secretion, androgen production and ovulatory function may support progression towards the adult PCOS phenotype.
Key words: androgens / endocrinology / gonadotrophin / polycystic ovaries / progesterone
|
Introduction |
|---|
In 1935, Stein and Leventhal first described the association of amenorrhoea, hirsutism, obesity and polycystic ovaries in a group of reproductive aged women (Stein and Leventhal, 1935
). Polycystic ovary syndrome (PCOS), as it is now known, is a common disorder, affecting approximately 6–8% of women during their childbearing years (Diamanti-Kandarakis et al., 1999; Asuncion et al., 2000; Azziz et al., 2004). Although it is a heterogeneous disorder with variable clinical presentations, it is defined clinically by ovulatory dysfunction and hyperandrogenism—with or without polycystic ovarian morphology—in the absence of other explanatory endocrinopathies (Zawadski and Dunaif, 1992). Ovulatory dysfunction generally manifests as oligomenorrhoea or amenorrhoea with consequent subfertility, and hyperandrogenism may be biochemical or clinical, with the latter including hirsutism, acne and androgenic alopoecia. PCOS is also commonly associated with obesity, insulin resistance and hyperinsulinaemia (Dunaif et al., 1989; Carmina et al., 1992; Legro et al., 1998b; DeUgarte et al., 2005), and women with PCOS are at increased risk of various metabolic disorders such as diabetes mellitus (Ehrmann et al., 1999; Legro et al., 1999) and the metabolic syndrome (Glueck et al., 2003; Apridonidze et al., 2005; Sam et al., 2005).
Despite decades of research, the aetiology of PCOS remains unclear. Altered ovarian steroidogenesis, hyperinsulinaemia and neuroendocrine abnormalities have all been proposed as primary aetiological factors (Yen et al., 1970; Dunaif and Graf, 1989; Poretsky and Piper, 1994; Ehrmann et al., 1995; Franks, 1995; Utiger, 1996; Gilling-Smith et al., 1997). A role for intrinsic abnormalities in ovarian steroidogenesis is supported by in vitro studies demonstrating that ovarian theca cells from women with PCOS produce excessive androgens (Gilling-Smith et al., 1994; Nelson et al., 1999). In addition, women with PCOS demonstrate abnormal ovarian steroid responses to gonadotropin stimulation (Ehrmann et al., 1995; Ibanez et al., 1996; Gilling-Smith et al., 1997; McCartney et al., 2004). Data also suggest an important role for insulin in the pathogenesis of PCOS (Dunaif, 1997; Poretsky et al., 1999). Insulin acts synergistically with LH to stimulate ovarian androgen production (Barbieri et al., 1986; Nestler et al., 1998a), and insulin suppresses hepatic production of sex-hormone-binding globulin, resulting in higher levels of free or bioavailable testosterone. Not only is there a high prevalence of insulin resistance and hyperinsulinaemia in women with PCOS (DeUgarte et al., 2005), but both hyperandrogenaemia and ovulatory function improve with treatments that decrease plasma insulin and/or improve insulin signalling (Dunaif, 1997; Poretsky et al., 1999; Azziz et al., 2001; Baillargeon et al., 2004). Neuroendocrine abnormalities also occur in PCOS and are clearly involved in the pathophysiology of this disorder. PCOS is marked by excessive LH pulsatility and relative FSH deficiency (Rebar et al., 1976; Waldstreicher et al., 1988; Taylor et al., 1997). In women with PCOS, treatment with GnRH antagonists results in an acute, dose-dependent reduction in both LH and testosterone levels (Hayes et al., 1998), and treatment with long-term GnRH agonists leads to suppression of ovarian androgen production to post-menopausal levels (Chang et al., 1983; Steingold et al., 1987). Thus, LH is the proximate physiological stimulus for androgen synthesis by ovarian theca cells and plays an important role in maintaining hyperandrogenaemia.
The initial manifestations of PCOS are frequently peripubertal in onset. Many women with PCOS develop clinical hyperandrogenism shortly after puberty, and the majority never establish regular menstrual periods. This suggests that PCOS has pre- or peripubertal origins. Many of the aforementioned abnormalities associated with PCOS, including aberrant neuroendocrine function, are also observed in adolescents with hyperandrogenaemia, which is felt to represent a forerunner of adult PCOS. The origin of elevated androgens in adult PCOS and adolescent hyperandrogenaemia is probably multifactorial, with entities such as abnormal ovarian steroidogenesis, hyperinsulinaemia and increased LH drive probably being complementary and synergistic, and with the relative contributions of each varying from individual to individual. Whether primary or secondary, neuroendocrine abnormalities make an important contribution to the pathogenesis and development of PCOS and will be the focus of this review.
|
Gonadotropin regulation during the ovulatory menstrual cycle |
|---|
 TOPAbstractIntroduction Gonadotropin regulation during... Gonadotropin secretion in PCOSAetiology of abnormal...Gonadotropin secretion in normal...A hypothesis regarding the...SummaryReferences |
|---|
The ovulatory menstrual cycle is dependent on a series of highly complex and integrated interactions between the central nervous system (CNS), hypothalamus, pituitary gland and ovaries (in addition to hormone-responsive elements such as the endometrium). However, the hypothalamo-pituitary unit is the primary driver of reproductive function. A group of functionally integrated hypothalamic neurons—known collectively as the GnRH pulse generator—secrete GnRH in a pulsatile fashion into the hypophyseal portal system. GnRH pulses stimulate synthesis and secretion of LH and FSH from pituitary gonadotropes. Interestingly, LH and FSH are produced by the same gonadotrope cell, yet plasma LH and FSH concentrations vary discordantly throughout the menstrual cycle. For instance, FSH exceeds LH in the early follicular phase, whereas LH predominates in the late follicular phase (Marshall and Kelch, 1986
).
The frequency of GnRH pulses in part determines which gonadotropin hormone is preferentially synthesized and secreted, with rapid GnRH pulses favouring LH, whereas slower GnRH pulses favour FSH (Gross et al., 1987; Spratt et al., 1987). For example, in vivo primate studies demonstrate that a GnRH pulse frequency of one pulse per hour favours secretion of LH relative to FSH; however, when pulse frequency is reduced to one pulse every 3 h, FSH concentration doubles whereas that of LH falls (Wildt et al., 1981). Rodent studies suggest a mechanism for this phenomenon. In rats, rapid GnRH pulses favour pituitary gonadotrope expression of the LH-ß gene, whereas slower pulses favour expression of the FSH-ß gene (Dalkin et al., 1989). This is in part explained by increased expression of pituitary follistatin induced by rapid GnRH pulses (Kirk et al., 1994); follistatin binds to the intragonadotrope activin, thereby inhibiting its stimulatory effect on FSH-ß expression and shifting the balance towards LH production.
Ovarian hormones influence GnRH pulse frequency. Importantly, elevations in progesterone levels result in slowing of LH and, by inference, GnRH pulse frequency. This slowing is seen during the luteal phase as well as after exogenous administration of physiologic doses of progesterone during the follicular phase (Soules et al., 1984). Estradiol (E2) plays a permissive role in progesterone suppression of GnRH pulsatility (Goodman et al., 1981; Nippoldt et al., 1989), probably through up-regulation of hypothalamic progesterone receptors (MacLusky and McEwen, 1978; Romano et al., 1989). This modulatory effect of ovarian hormones seems to be a primary regulator of GnRH pulse frequency, because the GnRH pulse generator appears to have an intrinsic firing frequency of approximately one pulse per hour after puberty. For example, in vitro studies reveal that the isolated adult human mediobasal hypothalamus releases pulses of GnRH with a periodicity of 60–100 min (Rasmussen et al., 1989). A similar pulse frequency is seen in women after natural and surgical menopause as well as in women with premature ovarian failure, all of whom lack ovarian steroid feedback (Rossmanith et al., 1990; Gill et al., 2002). Additionally, this rate is not exceeded at any point during the ovulatory menstrual cycle (Filicori et al., 1986; Rossmanith et al., 1990; Adams et al., 1994). Given this intrinsic pulse frequency, alterations in GnRH pulsatility in normally cycling women appear to be mediated through the imposition and removal of negative feedback, with progesterone being primary in this role.
The changes in GnRH pulse frequency and levels of gonadotropins, ovarian sex steroids and inhibins during the course of a normal ovulatory menstrual cycle are depicted in Figure 1, and the differential control of gonadotropin secretion may be understood in the context of the aforementioned data. GnRH pulsatility gradually increases during the follicular phase, apparently as a result of gradual loss of the restraining influence of progesterone (McCartney et al., 2002), achieving a peak frequency of approximately one pulse per hour in the late follicular phase (Filicori et al., 1986). This leads to an increase in LH and a decrease in FSH during the latter half of the follicular phase. In the late follicular phase, E2 production by the dominant follicle becomes sufficient to elicit marked increases in pituitary LH release in response to the rapid GnRH pulses resulting in positive feedback; this results in the mid-cycle LH surge, which in turn provokes ovulation (Karsch et al., 1973; Adams et al., 1994). Following ovulation, the corpus luteum produces E2 and progesterone, which again slow GnRH pulse frequency to one pulse every 3–4 h (Filicori et al., 1986), resulting in preferential synthesis and release of FSH and initiation of the next wave of follicular development.
|
Given the intricate and interdependent series of interactions that must take place for ovulatory cyclicity to occur, perturbations at any level of the hypothalamic–pituitary–ovarian axis can result in ovulatory dysfunction.
|
Gonadotropin secretion in PCOS |
|---|
Women with PCOS have consistent abnormalities in gonadotropin secretion. In a study that carefully excluded post-ovulatory subjects, 75% of women with PCOS had elevated LH levels and 94% had elevated LH : FSH ratios (Taylor et al., 1997
). PCOS is characterized by increased LH pulse amplitude and exaggerated LH responses to exogenous GnRH, whereas plasma FSH levels are relatively low (Rebar et al., 1976; Taylor, 1998; Marshall and Eagleson, 1999). Waldstreicher and colleagues demonstrated that women with PCOS have persistently rapid LH (GnRH) pulse frequency, in the order of one pulse per hour, without the normal cyclic variation seen in ovulatory women (Figure 2) (Waldstreicher et al., 1988). Such persistently rapid GnRH pulsation favours synthesis and secretion of LH over FSH, helping to explain the elevated LH levels and LH : FSH ratios characteristic of PCOS (Taylor et al., 1997). Because a frequency of approximately one pulse per hour appears to be inherent to the post-pubertal GnRH pulse generator, the persistently rapid LH pulsatility characteristic of PCOS probably signifies a failure of the systems necessary to suppress GnRH pulsatility, rather than representing an acceleration of the GnRH pulse generator. This failure of intermittent suppression may be the result of primary hypothalamic defects, an abnormal hormonal milieu or a combination of the two.
|
|
Aetiology of abnormal gonadotropin secretion in PCOS |
|---|
Hypothalamic neurotransmitters
In animals, dopamine and opioid pathways are inhibitory to GnRH neurons, whereas noradrenergic pathways are stimulatory. Data regarding the effects of 
-aminobutyric acid (GABA) have been mixed (DeFazio et al., 2002; Han et al., 2002, 2004). Information regarding CNS pathways in humans is by necessity indirect and generally derived through the use of medications that stimulate or inhibit these pathways. Decreased dopaminergic tone was hypothesized to play a role in PCOS, given that approximately 17% of women with PCOS have coincident moderate hyperprolactinaemia (Luciano et al., 1984). However, the dopamine agonist bromocriptine does not improve clinical or biochemical parameters in women with PCOS (Buvat et al., 1986; Murdoch et al., 1987). Decreased opioid tone was hypothesized to contribute to PCOS, as central opioidergic regulation mediates the normal luteal phase slowing of LH (GnRH) pulsatility (Quigley and Yen, 1980; Wardlaw et al., 1982; Soules et al., 1984), which is absent in PCOS. However, administration of exogenous progesterone to women with PCOS slows GnRH pulse frequency through an opiate-dependent process (Berga and Yen, 1989). Therefore, any apparent reduction of opioid activity in PCOS is probably secondary to the absence of progesterone effects (e.g. secondary to anovulation), rather than primary defects in opioid tone. Potential roles for noradrenergic and GABAergic pathways in PCOS have been explored via administration of the 
1-adrenoceptor antagonist thymoxamine (Paradisi et al., 1987) and valproate, which increases GABA concentrations (Popovic and Spremovic, 1995). Neither study documented changes in mean LH concentrations or LH pulse frequency (Paradisi et al., 1987; Popovic and Spremovic, 1995). Thus, to date, no primary hypothalamic defects have been clearly identified to explain the neuroendocrine abnormalities in PCOS. This raises the possibility that the neuroendocrine abnormalities are instead secondary to an abnormal hormonal environment.
An earlier theory, known as the ‘estrone hypothesis’, suggested that excess androstenedione was peripherally aromatized to estrone, which in turn caused increased LH secretion (Yen et al., 1976). This theory was based on the elevated estrone levels seen in PCOsS and the stimulatory effects of estrogens on LH secretion during the mid-cycle LH surge. However, administration of exogenous estrone does not increase basal or GnRH-stimulated LH concentrations in controls or in women with PCOS (Chang et al., 1982). In addition, use of the peripheral aromatase inhibitor testolactone to block conversion of androstenedione to estrone does not reduce LH pulse frequency (Dunaif et al., 1985). Therefore, available evidence does not support a role for estrone in the pathogenesis of PCOS.
Insulin resistance is a common metabolic abnormality seen in obese and, to a lesser degree, lean women with PCOS, with studies estimating a prevalence between 53 and 76% (Dunaif et al., 1989; Carmina et al., 1992; Legro et al., 1998b; DeUgarte et al., 2005). Although the majority of women are able to compensate for the insulin resistance with increased insulin secretion, 7–10% of reproductive age women with PCOS meet the criteria for type 2 diabetes mellitus (Ehrmann et al., 1999; Legro et al., 1999). Similarly, women with type 2 diabetes are over six times as likely as non-diabetic, age- and weight-matched controls to have PCOS (Tok et al., 2004). Insulin acts synergistically with LH to stimulate ovarian androgen production (Barbieri et al., 1986; Nestler et al., 1998a), and insulin suppresses hepatic production of sex-hormone-binding globulin (Dunkel et al., 1985), resulting in higher levels of free or bioavailable testosterone. Treatment of women with PCOS with either metformin or thiazoladinediones not only ameliorates insulin resistance but also results in improvements in hyperandrogenaemia and ovulatory function (Dunaif et al., 1996; Nestler and Jakubowicz, 1996; Ehrmann et al., 1997; Nestler et al., 1998b; Moghetti et al., 2000; Fleming et al., 2002; Ghazeeri et al., 2003; Lord et al., 2003). Although most of the androgen reduction related to metformin and thiazoladinedione treatment is thought to reflect decreases in circulating insulin levels, there is also in vitro evidence for direct effects on ovarian steroidogenesis (Attia et al., 2001; Mitwally et al., 2002; Mansfield et al., 2003).
Although hyperinsulinaemia appears to be an important contributor in the pathogenesis of PCOS, its role in neuroendocrine dysfunction is unclear. Insulin increases basal and GnRH-stimulated LH and FSH secretion from cultured rat pituitary cells in vitro (Adashi et al., 1981; Soldani et al., 1994). However, in analogous in vivo studies, insulin does not augment gonodotrope responses to GnRH (Poretsky et al., 1988). In women with PCOS, infusions of exogenous insulin do not alter LH secretion (Dunaif and Graf, 1989; Patel et al., 2003; Mehta et al., 2005), and reduction of insulin in PCOS has inconsistent effects on LH levels (Poretsky et al., 1999). Short-term treatment with metformin does not slow LH pulse frequency, despite significant improvements in insulin levels (Eagleson et al., 2003). Likewise, treatment with the thiazoladinedione pioglitazone does not alter LH pulse patterns, despite improvement in insulin sensitivity (Mehta et al., 2005). Moreover, LH appears to be negatively correlated with obesity (Morales et al., 1996; Taylor et al., 1997), despite a positive correlation between adiposity and insulin concentrations. Together, these findings suggest that the hyperinsulinaemia associated with PCOS does not directly result in neuroendocrine abnormalities. However, hyperinsulinaemia may bring about important changes in hypothalamic function indirectly by increasing androgen levels (discussed below).
Cyclic (luteal) increases in progesterone result in regular, periodic slowing of LH (GnRH) pulsatility in ovulatory women. However, a defining feature of PCOS is chronic oligoovulation or anovulation. Therefore, women with PCOS do not regularly experience the post-ovulatory rise in progesterone seen in normally cycling women. Undoubtedly, the lower progesterone levels associated with anovulation play a role in the persistently rapid GnRH pulse frequency characteristic of PCOS in adult women. However, this alone does not appear to be a sufficient explanation for the neuroendocrine abnormalities of PCOS. Normally cycling women have occasional anovulatory cycles, and many women with PCOS ovulate intermittently, albeit infrequently. The modest rise in E2 and progesterone levels seen in anovulatory cycles in normal women results in the slowing of LH pulsatility that may be equivalent to that seen during an ovulatory cycle (Clayton et al., 1987). Conversely, intermittent ovulation does not correct the rapid LH pulse frequency in PCOS, and LH levels are again elevated some 10–14 days after progesterone levels fall (Taylor et al., 1997). Also, adolescent girls with hyperandrogenaemia—which is felt to represent a forerunner of PCOS—have more rapid LH (GnRH) pulse frequency than normal controls, even before menarche when ovulatory cycles have not yet been established (Apter et al., 1994).
Another possible contributor to decreased progesterone effects in PCOS is reduced sensitivity of the GnRH pulse generator to the negative feedback effects of sex steroids. To determine whether the increased GnRH pulsatility in PCOS is explained simply by abnormal sex steroid levels in the setting of chronic anovulation, Daniels and Berga, in their study, used oral contraceptives to simulate luteal phase E2 and progesterone concentrations in women with PCOS (Daniels and Berga, 1997). Oral contraceptives slowed LH pulse frequency in hyperandrogenaemic anovulatory women to some extent, but not to the same degree seen in normal controls. Subsequent studies using 7 days of E2 and progesterone confirmed that women with PCOS have reduced hypothalamic sensitivity to progesterone-mediated suppression of LH (GnRH) pulse frequency compared with normal women (Pastor et al., 1998). Thus, women with PCOS require higher levels of progesterone to achieve the same degree of GnRH suppression as ovulatory controls. Interestingly, similar findings are observed in some adolescents with hyperandrogenaemia (Chhabra et al., 2005), which may help explain the emergence of neuroendocrine abnormalities during puberty. Progesterone sensitivity is restored in adult PCOS with the use of the androgen receptor blocker flutamide, indicating that reduced progesterone sensitivity is secondary to hyperandrogenaemia, rather than a primary hypothalamic defect (Figure 3) (Eagleson et al., 2000).
|
Hyperandrogenism, like chronic anovulation, is a hallmark of PCOS, and its origin appears to be multifactorial, with increased LH and insulin stimulation of ovaries that appear to hypersecrete androgens (Chang et al., 1983; Barbieri et al., 1986; Steingold et al., 1987; Gilling-Smith et al., 1994; Nestler et al., 1998a; Nelson et al., 1999). Regardless of the aetiology, elevated androgen levels appear to have important neuroendocrine consequences.
Early in vitro studies found that androgens increase GnRH pulse frequency in isolated rat GnRH neurons (Melrose and Gross, 1987). Initial studies in women with PCOS found that neither androgen infusions (Dunaif, 1986) nor short-term (less than 1 month) treatments with the androgen receptor blocker flutamide (Sir-Petermann et al., 1993; Eagleson et al., 2000) change basal LH concentrations or pulsatility. However, as discussed above, short-term treatment with flutamide restores hypothalamic sensitivity to progesterone-mediated slowing of LH pulsatility in women with PCOS (Eagleson et al., 2000), and long-term (6 months) administration of flutamide may ameliorate many of the reproductive manifestations of PCOS (De Leo et al., 1998).
Prenatal androgen exposure has been used as an animal model of PCOS, with some interesting and informative findings. Prenatally androgenized female monkeys, sheep and rats exhibit abnormal neuroendocrine function, such as increased LH concentrations and LH (GnRH) pulse frequency during subsequent pubertal maturation (Dumesic et al., 1997; Robinson et al., 1999; Abbott et al., 2005; Foecking et al., 2005). Similar to women with PCOS, prenatally androgenized sheep demonstrate decreased sensitivity to progesterone suppression of LH (GnRH) pulse frequency (Robinson et al., 1999). Foecking and co-workers recently demonstrated that prenatally androgenized rats have decreased basal and E2-induced progesterone receptors in the hypothalamus, suggesting a possible mechanism for androgen-mediated hypothalamic progesterone insensitivity (Foecking et al., 2005). Prenatal androgenization also increases GABAergic drive to GnRH neurons in mice, an abnormality which can be reversed with flutamide treatment as in adults (Sullivan and Moenter, 2004).
In addition to providing a useful model for mechanistic studies that are difficult to carry out in humans, there is some evidence that prenatal androgenization may play a role in human disease. Women with congenital adrenal hyperplasia have persistent evidence of ovarian hyperandrogenism even when well controlled on glucocorticoid treatment, whereas their counterparts with late-onset 21-hydroxylase deficiency do not (Barnes et al., 1994). Women with PCOS may maintain elevated androgen levels during pregnancy (Sir-Petermann et al., 2002), thus potentially exposing their unborn daughters to elevated androgen levels in utero. PCOS clusters in families (Legro et al., 1998a; Kahsar-Miller et al., 2001), and the search for specific genetic causes has proven elusive. In theory, prenatal exposure to maternal hyperandrogenaemia may in part explain the hereditary nature of the syndrome. However, placental aromatase is highly effective in converting androgens to estrogens, typically protecting the fetus from excess androgen exposure in utero. In addition, the doses of androgens given in animal models create androgen levels that far exceed those seen in pregnant women with PCOS. Therefore, it is unclear if the more subtle elevations in androgens, typical of PCOS, reach the fetus or have any effect on development.
As described above, elevated androgens have significant neuroendocrine effects, including impairment of hypothalamic progesterone sensitivity. This relative inability to slow GnRH pulse frequency in part explains the persistently rapid GnRH pulse frequency in PCOS. This defect promotes excessive LH feedforward drive with subsequent excessive ovarian androgen production as well as relative FSH deficiency that contributes to ovulatory dysfunction. As discussed above, similar neuroendocrine abnormalities are observed in adolescent hyperandrogenaemia, yet it is unclear how these abnormalities are established during puberty. We have proposed one hypothetical paradigm that may explain the genesis of neuroendocrine abnormalities in adolescents with hyperandrogenaemia. Below, we discuss the reproductive neuroendocrinology of normal puberty together with neuroendocrine findings in adolescents with hyperandrogenaemia.
|
Gonadotropin secretion in normal female puberty and in adolescents girls with hyperandrogenaemia |
|---|
|
TOPAbstractIntroductionGonadotropin regulation during...Gonadotropin secretion in PCOSAetiology of abnormal...Gonadotropin secretion in normal...A hypothesis regarding the...SummaryReferences |
|---|
The GnRH pulse generator is active during fetal development and the early neonatal period, but the GnRH secretory system becomes relatively quiescent during the 6–9 months after birth and remains so during childhood. This juvenile period is characterized by low levels of LH and FSH with a high FSH : LH ratio. GnRH pulse frequency is slow (i.e. one pulse every 4–6 h in prepubertal girls), and LH pulse amplitude is very low (Cemeroglu et al., 1996
; Clark et al., 1997). The onset of puberty is heralded by nocturnal, sleep-entrained increases in LH pulse amplitude and frequency. This results in a diurnal pattern of LH pulsatility that precedes physical manifestations of pubertal maturation by approximately 2 years (Boyar et al., 1972; Wu et al., 1989; Wennink et al., 1990). Nocturnal increases in LH secretion are followed by early morning rises in sex steroids such as E2 and testosterone (Norjavaara et al., 1996; Ankarberg and Norjavaara, 1999; Mitamura et al., 2000). As puberty progresses, daytime LH pulsatility increases, rendering the diurnal changes less prominent; by late puberty, sleep-associated increases in LH pulsatility are lost (Boyar et al., 1972; Wennink et al., 1990).
The cause of diurnal changes in LH pulsatility during puberty is unclear. It could primarily reflect diurnal feedforward drive from higher CNS inputs. Indeed, a role for higher CNS inputs is suggested by the observation that increases in LH pulsatility closely follow the onset of sleep (Boyar et al., 1972). However, diurnal changes in sex steroids may also influence diurnal changes in LH pulsatility via intermittent negative feedback or by facilitating changes in higher CNS functions. This possibility is suggested by several observations. The early morning peaks of sex steroid concentrations precede the time when LH pulsatility wanes. Also, in contrast to the normal patterns of diurnal LH and E2 secretion in early pubertal girls, age-matched girls with gonadal dysgenesis appear to demonstrate elevated LH pulse frequencies at all time periods (Cemeroglu et al., 1996), suggesting that an ovarian factor (e.g. E2) may be involved in the day–night differences of pubertal LH pulse frequency. E2 infusion in peripubertal girls (Cemeroglu et al., 1998) diminishes nocturnal increases in LH pulsatility. While data in girls are lacking, infusions of testosterone in boys also suppress nocturnal LH (GnRH) pulsatility (Kletter et al., 1994), suggesting that the nocturnal rise in testosterone may also play a role in girls. In addition, a role for sex steroids in the diurnal pattern of GnRH secretion is in keeping with the notion (described above) that changes in GnRH pulse frequency are determined by the application and removal of negative feedback.
The mechanisms by which the hypothalamic–pituitary–gonadal axis is quiescent during childhood and reactivates during puberty also remain unclear. In girls, LH pulse frequency and amplitude increases four- and nine-fold across pubertal maturation (Apter et al., 1993). The relative dormancy of this axis during childhood may reflect inhibition by higher CNS pathways. However, the hypothalamus also appears to be exquisitely sensitive to sex steroid feedback during childhood, and as puberty advances, there is decreasing sensitivity to E2 inhibition of LH secretion (Kelch et al., 1973; Rapisarda et al., 1983). Thus, the gradual loss of sensitivity to feedback inhibition contributes to the increase in LH pulsatility during puberty. The cause of decreasing sensitivity to feedback inhibition is unknown, but given that hyperandrogenaemia appears to account for the abnormal reduction of feedback sensitivity in adults with PCOS, one possibility is that the gradual increase in androgen concentrations characteristic of normal female puberty (Ankarberg and Norjavaara, 1999; Mitamura et al., 2000) may mediate the reduction in feedback sensitivity associated with pubertal progression.
Hyperandrogenism during adolescence is thought to represent a forerunner of PCOS. A study of girls with menstrual irregularities showed that although some subjects normalize endocrine function as they mature, the majority maintain hyperandrogenaemia, elevated LH levels and polycystic ovaries, characteristic of PCOS (Venturoli et al., 1987). Additionally, adolescent hyperandrogenaemia is associated with higher androgen levels and lower fertility rates in adulthood (Apter and Vihko, 1990).
Relatively little is known about changes in gonadotropin secretion in adolescents with hyperandrogenism. Many such adolescents demonstrate elevated mean LH, LH pulse frequency and LH pulse amplitude (Venturoli et al., 1992). Compared with age-matched controls, girls with hyperandrogenism exhibit higher daytime and nighttime LH pulse frequency and higher mean 24 h LH : FSH ratios (Apter et al., 1994). The transition from primarily sleep-related LH secretion to predominance of daytime LH secretion is advanced by some 2 years in hyperandrogenaemic girls (Apter et al., 1994). Another study revealed that four of five adolescents with hyperandrogenaemia demonstrated abnormal LH pulsatility with increased LH pulsatility during the day, which contrasts with sleep-associated increases, characteristic of normal adolescent controls (Zumoff et al., 1983). Also, some hyperandrogenaemic adolescents exhibit relative GnRH pulse generator resistance to the feedback actions of progesterone and E2 (Chhabra et al., 2005). Overall, these data suggest that neuroendocrine (LH) abnormalities are present during puberty in adolescents who will go on to develop adult PCOS.
|
A hypothesis regarding the genesis of neuroendocrine dysfunction in adolescents with hyperandrogenaemia |
|---|
The cause(s) of neuroendocrine abnormalities in adolescents with hyperandrogenism is (are) unknown, but hyperandrogenaemia itself may play a prominent role. Girls with precocious adrenarche are at increased risk of developing PCOS (Ibanez et al., 2000
). Women with congenital adrenal hyperplasia demonstrate elevated LH (Barnes et al., 1994), as do prenatally androgenized animals (Dumesic et al., 1997; Robinson et al., 1999). The mechanism accounting for this association is unclear, but, by analogy, to what is observed in adults with PCOS, we have proposed that neuroendocrine abnormalities are partly related to androgen-dependent decreases in GnRH pulse generator sensitivity to the negative feedback actions of ovarian steroids. Thus, it remains possible that any cause of hyperandrogenaemia—e.g. hyperandrogenaemia during fetal life or associated with premature adrenarche (Ibanez et al., 2000), obesity (Wabitsch et al., 1995; Reinehr et al., 2005), hyperinsulinaemia (Dunaif, 1997; Poretsky et al., 1999), polycystic ovarian morphology (Adams et al., 2004) or inherent abnormalities of ovarian steroidogenesis (Ehrmann et al., 1995)—may produce abnormalities in feedback control of pulsatile GnRH secretion, leading to persistently fast GnRH pulses. Of course, the resulting abnormalities in gonadotropin secretion (increased LH and relative decreases in FSH) would further contribute to ovarian hyperandrogenaemia and ovulatory dysfunction, establishing a vicious circle that promotes the progression towards the adult PCOS phenotype. Although the importance of diurnal GnRH pulsatility during normal puberty is not established, we suggest that diurnal (daytime) slowing of GnRH pulse frequency is important to support pituitary FSH secretion during puberty. We hypothesize that early morning increases in gonadal steroids normally reduce GnRH and LH pulsatility during the following day. Inasmuch as the diurnal increases in sex steroids are important in mediating diurnal changes in GnRH pulsatility, interference with negative feedback of the GnRH pulse generator would be expected to diminish diurnal changes in pulse frequency. If hyperandrogenaemia results in relative GnRH pulse generator resistance to negative feedback suppression in peripubertal girls (as it does in adults), then any cause of hyperandrogenaemia during puberty may produce elevated GnRH pulse frequency throughout the day and night, increasing LH and decreasing FSH synthesis and secretion, thereby contributing to both hyperandrogenaemia and ovulatory dysfunction. Again, initial hyperandrogenaemia could occur via a number of mechanisms including obesity, hyperinsulinaemia, abnormal steroidogenesis, etc. However, the concomitant disruption of neuroendocrine function would serve to further exacerbate hyperandrogenaemia, supporting the progression towards the adult PCOS phenotype.
Additionally, if testosterone modulation of hypothalamic E2 and progesterone sensitivity indeed play a role in mediating gradual decreases of sensitivity feedback during normal puberty, then it is logical that girls with hyperandrogenaemia would progress to an adult pattern at an earlier age, and thereafter to exaggerated LH pulsatility (Figures 4 and 5). This could in part explain why obesity is associated with early sexual maturation (Wang, 2002): peripubertal obesity is associated with hyperandrogenaemia (Reinehr et al., 2005), which may result in an early decrease in hypothalamic sensitivity to negative feedback, permitting advanced development of LH pulsatility and subsequent puberty.
|
|
Clearly, this theoretical paradigm requires further study, but it is consistent with available data, and it offers a coherent explanation for what has been observed in adult and adolescent PCOS. Further investigations are needed to explore the origins and implications of adolescent hyperandrogenaemia. The link between obesity and elevated androgen levels is especially relevant in the light of the rising incidence of childhood obesity in the developed world (Ogden et al., 2002; Hedley et al., 2004). The International Obesity Task Force estimates that there are 30–45 million obese school-age children worldwide (Lobstein et al., 2004). Given that obesity is associated with hyperandrogenaemia, and adolescent hyperandrogenaemia often progresses to adult PCOS, the epidemic of childhood obesity may be followed by a rise in the incidence of PCOS. However, not all obese girls are hyperandrogenaemic, and not all hyperandrogenaemic girls experience adverse neuroendocrine effects (Chhabra et al., 2005). Further study is needed to determine what modulating factors determine which girls are negatively affected by pubertal hyperandrogenaemia. Elucidation of inciting factors will eventually lead to effective early intervention in susceptible populations. If the above concepts are correct, interruption of the cycle with anti-androgens or cyclic progesterone may possibly prevent the development of PCOS.
|
Summary |
|---|
PCOS remains an enigmatic disorder with variable clinical presentations and uncertain aetiology. Neuroendocrine abnormalities, including rapid LH (GnRH) pulse frequency, elevated plasma LH and elevated LH : FSH ratios, are prominent in both adult PCOS and adolescent hyperandrogenaemia. Elevated androgens impair hypothalamic sensitivity to progesterone suppression of GnRH pulse frequency in women with PCOS. This hypothalamic progesterone insensitivity contributes to abnormal GnRH and gonadotropin secretion, impaired follicular development and increased ovarian androgen production. A subset of hyperandrogenaemic adolescent girls demonstrates similarly impaired hypothalamic progesterone sensitivity. Therefore, it is possible that peripubertal hyperandrogenaemia—whether it results from obesity, hyperinsulinaemia or abnormal adrenal or ovarian steroidogenesis—initiates a vicious cycle in which hyperandrogenaemia leads to neuroendocrine abnormalities which in turn perpetuate hyperandrogenaemia. Further studies are needed to explore these hypotheses and the origins of adolescent hyperandrogenaemia.
|
Acknowledgements |
|---|
We are grateful to S. Chhabra, MD, C.A. Eagleson, MD, C.L. Pastor, MD and K.A. Prendergast, MD for their previous contributions as clinical research fellows; A. Bellows, PhD, C. Chopra and Q.J.L. Okonkwo, MD for research coordination; the GCRC staff and nurses; and the Ligand Core Lab of the Center for Research in Reproduction. This work was supported by National Institutes of Health Grants RO1-HD-34179 and RO1-HD-33039 (J.C.M.), T32-HD-07382 (S.K.B.) and 1-K23-HD-044742 (C.R.M.); National Institute of Child Health and Human Development, National Institutes of Health, Specialized Cooperative Centers Program in Reproduction Research Grant U54-HD-28934 (J.C.M.); and GCRC Grant 5-MO1-RR-00847.
|
References |
|---|
Abbott DH, Barnett DK, Bruns CM and Dumesic DA (2005) Androgen excess fetal programming of female reproduction: a developmental aetiology for polycystic ovary syndrome? Hum Reprod Update 11,357–374.
Adams JM, Taylor AE, Schoenfeld DA, Crowley WF Jr and Hall JE (1994) The midcycle gonadotropin surge in normal women occurs in the face of an unchanging gonadotropin-releasing hormone pulse frequency. J Clin Endocrinol Metab 79,858–864.[Abstract]
Adams JM, Taylor AE, Crowley WF Jr and Hall JE (2004) Polycystic ovarian morphology with regular ovulatory cycles: insights into the pathophysiology of polycystic ovarian syndrome. J Clin Endocrinol Metab 89,4343–4350.
Adashi EY, Hsueh AJ and Yen SS (1981) Insulin enhancement of luteinizing hormone and follicle-stimulating hormone release by cultured pituitary cells. Endocrinology 108,1441–1449.
Ankarberg C and Norjavaara E (1999) Diurnal rhythm of testosterone secretion before and throughout puberty in healthy girls: correlation with 17beta-estradiol and dehydroepiandrosterone sulfate. J Clin Endocrinol Metab 84,975–984.
Apridonidze T, Essah PA, Iuorno MJ and Nestler JE (2005) Prevalence and characteristics of the metabolic syndrome in women with polycystic ovary syndrome. J Clin Endocrinol Metab 90,1929–1935.
Apter D and Vihko R (1990) Endocrine determinants of fertility: serum androgen concentrations during follow-up of adolescents into the third decade of life. J Clin Endocrinol Metab 71,970–974.
Apter D, Butzow TL, Laughlin GA and Yen SS (1993) Gonadotropin-releasing hormone pulse generator activity during pubertal transition in girls: pulsatile and diurnal patterns of circulating gonadotropins. J Clin Endocrinol Metab 76,940–949.[Abstract]
Apter D, Butzow T, Laughlin GA and Yen SS (1994) Accelerated 24-hour luteinizing hormone pulsatile activity in adolescent girls with ovarian hyperandrogenism: relevance to the developmental phase of polycystic ovarian syndrome. J Clin Endocrinol Metab 79,119–125.[Abstract]
Asuncion M, Calvo RM, San Millan JL, Sancho J, Avila S and Escobar-Morreale HF (2000) A prospective study of the prevalence of the polycystic ovary syndrome in unselected Caucasian women from Spain. J Clin Endocrinol Metab 85,2434–2438.
Attia GR, Rainey WE and Carr BR (2001) Metformin directly inhibits androgen production in human thecal cells. Fertil Steril 76,517–524.[CrossRef][Web of Science][Medline]
Azziz R, Ehrmann D, Legro RS, Whitcomb RW, Hanley R, Fereshetian AG, O’Keefe M and Ghazzi MN (2001) Troglitazone improves ovulation and hirsutism in the polycystic ovary syndrome: a multicenter, double blind, placebo-controlled trial. J Clin Endocrinol Metab 86,1626–1632.
Azziz R, Woods KS, Reyna R, Key TJ, Knochenhauer ES and Yildiz BO (2004) The prevalence and features of the polycystic ovary syndrome in an unselected population. J Clin Endocrinol Metab 89,2745–2749.
Baillargeon JP, Jakubowicz DJ, Iuorno MJ, Jakubowicz S and Nestler JE (2004) Effects of metformin and rosiglitazone, alone and in combination, in nonobese women with polycystic ovary syndrome and normal indices of insulin sensitivity. Fertil Steril 82,893–902.[CrossRef][Web of Science][Medline]
Barbieri RL, Makris A, Randall RW, Daniels G, Kistner RW and Ryan KJ (1986) Insulin stimulates androgen accumulation in incubations of ovarian stroma obtained from women with hyperandrogenism. J Clin Endocrinol Metab 62,904–910.
Barnes RB, Rosenfield RL, Ehrmann DA, Cara JF, Cuttler L, Levitsky LL and Rosenthal IM (1994) Ovarian hyperandrogynism as a result of congenital adrenal virilizing disorders: evidence for perinatal masculinization of neuroendocrine function in women. J Clin Endocrinol Metab 79,1328–1333.[Abstract]
Berga SL and Yen SS (1989) Opioidergic regulation of LH pulsatility in women with polycystic ovary syndrome. Clin Endocrinol (Oxf) 30,177–184.[Medline]
Boyar R, Finkelstein J, Roffwarg H, Kapen S, Weitzman E and Hellman L (1972) Synchronization of augmented luteinizing hormone secretion with sleep during puberty. N Engl J Med 287,582–586.[Web of Science][Medline]
Buvat J, Buvat-Herbaut M, Marcolin G, Racadot A, Fourlinnie JC, Beuscart R and Fossati P (1986) A double blind controlled study of the hormonal and clinical effects of bromocriptine in the polycystic ovary syndrome. J Clin Endocrinol Metab 63,119–124.
Carmina E, Koyama T, Chang L, Stanczyk FZ and Lobo RA (1992) Does ethnicity influence the prevalence of adrenal hyperandrogenism and insulin resistance in polycystic ovary syndrome? Am J Obstet Gynecol 167,1807–1812.[Web of Science][Medline]
Cemeroglu AP, Foster CM, Warner R, Kletter GB, Marshall JC and Kelch RP (1996) Comparison of the neuroendocrine control of pubertal maturation in girls and boys with spontaneous puberty and in hypogonadal girls. J Clin Endocrinol Metab 81,4352–4357.[Abstract]
Cemeroglu AP, Kletter GB, Guo W, Brown MB, Kelch RP, Marshall JC, Padmanabhan V and Foster CM (1998) In pubertal girls, naloxone fails to reverse the suppression of luteinizing hormone secretion by estradiol. J Clin Endocrinol Metab 83,3501–3506.
Chang RJ, Mandel FP, Lu JK and Judd HL (1982) Enhanced disparity of gonadotropin secretion by estrone in women with polycystic ovarian disease. J Clin Endocrinol Metab 54,490–494.
Chang RJ, Laufer LR, Meldrum DR, DeFazio J, Lu JK, Vale WW, Rivier JE and Judd HL (1983) Steroid secretion in polycystic ovarian disease after ovarian suppression by a long-acting gonadotropin-releasing hormone agonist. J Clin Endocrinol Metab 56,897–903.
Chhabra S, McCartney CR, Yoo RY, Eagleson CA, Chang RJ and Marshall JC (2005) Progesterone inhibition of the hypothalamic gonadotropin-releasing hormone pulse generator: evidence for varied effects in hyperandrogenemic adolescent girls. J Clin Endocrinol Metab 90,2810–2815.
Clark PA, Iranmanesh A, Veldhuis JD and Rogol AD (1997) Comparison of pulsatile luteinizing hormone secretion between prepubertal children and young adults: evidence for a mass/amplitude-dependent difference without gender or day/night contrasts. J Clin Endocrinol Metab 82,2950–2955.
Clayton RN, Royston JP, Chapman J, Wilson M, Obhrai M, Sawers RS and Lynch SS (1987) Is changing hypothalamic activity important for control of ovulation? Br Med J (Clin Res Ed) 295,7–12.
Dalkin AC, Haisenleder DJ, Ortolano GA, Ellis TR and Marshall JC (1989) The frequency of gonadotropin-releasing-hormone stimulation differentially regulates gonadotropin subunit messenger ribonucleic acid expression. Endocrinology 125,917–924.
Daniels TL and Berga SL (1997) Resistance of gonadotropin releasing hormone drive to sex steroid-induced suppression in hyperandrogenic anovulation. J Clin Endocrinol Metab 82,4179–4183.
De Leo V, Lanzetta D, D’Antona D, la Marca A and Morgante G (1998) Hormonal effects of flutamide in young women with polycystic ovary syndrome. J Clin Endocrinol Metab 83,99–102.
DeFazio RA, Heger S, Ojeda SR and Moenter SM (2002) Activation of A-type gamma-aminobutyric acid receptors excites gonadotropin-releasing hormone neurons. Mol Endocrinol 16,2872–2891.
DeUgarte CM, Bartolucci AA and Azziz R (2005) Prevalence of insulin resistance in the polycystic ovary syndrome using the homeostasis model assessment. Fertil Steril 83,1454–1460.[CrossRef][Web of Science][Medline]
Diamanti-Kandarakis E, Kouli CR, Bergiele AT, Filandra FA, Tsianateli TC, Spina GG, Zapanti ED and Bartzis MI (1999) A survey of the polycystic ovary syndrome in the Greek island of Lesbos: hormonal and metabolic profile. J Clin Endocrinol Metab 84,4006–4011.
Dumesic DA, Abbott DH, Eisner JR and Goy RW (1997) Prenatal exposure of female rhesus monke